[GSoC Project Proposal]: Echoshader: a package for interactive visualisation and dashboarding of ocean sonar data

[LOCEANlloydizard]

Project Description

Ocean sonar systems, such as echosounders, are a core tool for observing marine life, providing unprecedented spatial, temporal, and vertical resolution of animal distributions and behaviour. These systems are now widely deployed on ships, autonomous vehicles, and moorings, generating large and complex multivariate datasets that require advanced processing before they can be scientifically interpreted.

This project builds on Echopype, which standardises and processes sonar data from a wide range of platforms. The goal of this GSoC project is to further develop Echoshader, an open-source companion visualisation tool written in Python, originally developed by a GSoC’22 contributor to facilitate ocean sonar data visualisation; it already provides building blocks such as echogram plots, ship-track maps, and 3D curtain views.

We aim to further develop the Echoshader package by extending the current functionalities, and demonstrate its capabilities on multiple types of datasets (from ships, moorings, and autonomous vehicles) through robust cloud deployment. In doing so, we aim to integrate recent developments from engine_echo_data_viz into echoshader.

Expected Outcomes

Software architecture enhancement:

• An updated codebase compatible with the latest dependency versions.
• A unified widget system that can be used both in notebooks and in a web-based dashboard interface.
• A modular, robust, and maintainable echoshader architecture aligned with ongoing developments in Echopype.

New functionalities:

• Incorporate new features prototyped in Jupyter notebooks into Echoshader, including interactive annotation modification and 3D echogram visualization.
• Prototype and implement new features in Echoshader to support visualization of mooring data.

Demonstration and deployment:

• Demonstrations of Echoshader functionalities on several real-world datasets spanning different platforms.
• A deployable application to enable straightforward installation and cloud or local deployment for the acoustics and broader ocean sciences community.

Skills Required

A keen interest in environmental data, including oceanographic, acoustic, and geospatial data; Fluency in Python programming ; Experience or strong interest in working with large, multi-dimensional datasets and knowledge of xarray, dask, HoloViz, Plotly

Expected Project Size

175 hours

Project Difficulty

Intermediate

Mentor Name(s)

Lloyd Izard, Wu-Jung Lee, Valentina Staneva

Mentor GitHub Handle(s)

@LOCEANlloydizard, @leewujung, @valentina-s

Mentor Contact Email(s)

lloyd.izard@gmail.com, leewj@uw.edu, vms16@uw.edu

Contributor Evaluation/GSoC GitHub Issue Links

OSOceanAcoustics/echoshader#218

Additional Project Background/Links

• Lee, W.J., Staneva, V., Setiawan, L. D., Mayorga, E., Tuguinay, C., Butala, S., Lucca, B., & Lei, D. (2024). Echostack: A flexible and scalable open-source software suite for echosounder data processing. Proceedings of the 23rd Python in Science Conference (SciPy 2024).

• Lee, W.J., Setiawan, L., Tuguinay, C., Mayorga, E. and Staneva, V., 2024. Interoperable and scalable echosounder data processing with Echopype. ICES Journal of Marine Science, 81(10), pp.1941-1951.

• https://github.com/OSOceanAcoustics/echopype

• https://github.com/OSOceanAcoustics/echopype-examples

• https://github.com/OSOceanAcoustics/echoshader

• https://github.com/uw-echospace/engine_echo_data_viz

Previous Project Idea Link(s)

#35

[Safwannn89]

Hi @LOCEANlloydizard,
I’m Safwan, a 4th-year Data Science student. I’ve read all the details and requirements of the project, and I found it very interesting, especially the opportunity to contribute to ocean exploration. I would love to work on this project if possible. I understand that the project is not finalized yet, but in the meantime, could you please guide me on how I can start working on some issues?

[LOCEANlloydizard]

Hi @Safwannn89, thank you for your interest!

To get started, we suggest setting up a good development environment, forking the repository, and having a look at its architecture. There is also documentation here. Then try running one of the notebooks?

Once you’re set, this issue is a good place to start and to communicate 🙂

[Safwannn89]

Hii @LOCEANlloydizard
I have gone through the documentation, and it was very exciting to explore these concepts. Following that, I have submitted the Pull Request for the environment modernization and notebook verification task.
I unpinned the NumPy and Zarr dependencies as requested and confirmed that everything is working correctly with NumPy 1.26.4 and Zarr 2.16.1.

The PR is ready for your review whenever you have a moment!
PR #221.

[LOCEANlloydizard]

Hi @Safwannn89,

sorry for the delay in my reply! I’ve seen your PR, that’s great, thnk you for your contribution! I’ll have a look as soon as possible, most likely next week.

In the meantime, I just wanted to make sure you’re aware that the organization selections for GSoC haven’t happened yet. It’s generally recommended to wait until we know whether we’re selected as a GSoC organization before working further on these projects. We will let you know if we are selected!

Cheers!

[Safwannn89]

Thanks for your response @LOCEANlloydizard